Predicting Driver Route Choices from Delay Information

Introduction:

Traffic congestion poses a significant economic burden, with annual costs in the United States estimated to exceed $63 billion in wasted fuel and time, plus an additional $60 billion in lost productivity. To manage this, states increasingly utilize real-time traveler information systems, such as Dynamic Message Signs (DMS) and 511 systems, to encourage drivers to divert to alternative routes. This study investigated how effectively these systems influence driver behavior, specifically examining the impact of numerical versus categorical delay information, driver age, familiarity with alternative routes, and prior knowledge obtained via 511 systems. The researchers hypothesized that longer reported delays would increase diversion rates and that older adults might be more prone to divert due to past negative experiences with traffic (the “hot stove effect”) or benefit more from prior information due to reduced cognitive load.

Methodology:

The study utilized a midrange driving simulator featuring a full-sized vehicle and a 135° field of view to replicate a realistic interstate environment. The 48 participants were divided into two age cohorts: younger/middle-aged drivers (26–55 years) and older drivers (65–85 years), all of whom were active licensed drivers with at least 10 years of experience. Participants were subjected to a mixed experimental design where they encountered seven distinct scenarios featuring different delay messages: two categorical signs (e.g., “Delays Ahead”) and five numerical signs reporting specific delays ranging from 20 to 70 minutes. Familiarity was manipulated by telling some participants the travel times for alternative routes (familiar condition) while withholding this for others (unfamiliar condition). To test the impact of 511 systems, half of the participants received delay information at the start of their drive (prior knowledge), while the other half received it only via the DMS during the drive. Data were analyzed using a mixed between-within analysis of variance (ANOVA) to determine the significance of each factor on diversion frequencies.

Results:

The research demonstrated that numerical delay length is the most reliable predictor of route diversion, with higher delays significantly increasing the likelihood of drivers choosing an alternative route regardless of age or familiarity. While age did not have a significant main effect on numerical delays, a notable interaction occurred: older adults were much more likely to divert at shorter reported delays (20–30 minutes) compared to younger drivers, potentially because they perceive higher variability and risk in delay reports. For categorical signs, older drivers diverted significantly more often (21.39% probability) than younger drivers (7.95% probability). Furthermore, familiarity played a crucial role; when alternative route travel times were unknown, younger drivers almost never diverted, whereas older drivers diverted 28% of the time. However, when alternative times were provided, both age groups behaved similarly. Surprisingly, 511 prior information had no measurable effect on diversion choices for any group, suggesting that drivers may effectively process necessary information directly from road signs during transit. These findings suggest that traffic engineers can improve congestion management by providing specific alternative route travel times alongside main route delay information.

References

1. Kwon, J., Mauch, M., & Varaiya, P. (2006). Components of congestion: Delay from incidents, special events, lane closures, weather, potential ramp metering gain, and excess demand. Transportation Research Record: Journal of the Transportation Research Board, 1959, 84-91.
2. Public Works Magazine. (2005, July 1). U.S. traffic delays skyrocket.
3. Bonsall, P., & Parry, T. (1991). Using an interactive route-choice simulator to investigate drivers’ compliance with route guidance advice. Transportation Research Record, 1306, 59-68.
4. Ben-Elia, E. E. (2008). The combined effect of information and experience on drivers’ route-choice behavior. Transportation, 35(2), 165–177.
5. Abdel-Aty, M. A. (2004). Modeling drivers’ diversion from normal routes under ATIS using generalized estimating equations and binomial probit link function. Transportation, 31(3), 327–348.
6. Katsikopoulos, K. V., & Fisher, D. L. (2002). Risk attitude reversals in drivers’ route choice when range of travel time information is provided. Human Factors, 44(3), 466–473.
7. Katsikopoulos, K. V., & Fisher, D. L. (2000). The framing of drivers’ route choices when travel time information is provided under varying degrees of cognitive load. Human Factors, 42(3), 470–481.
8. Simon, H. A. (1957). Models of man. John Wiley & Sons, Inc.
9. Ben-Akiva, M. K. (1997). Simulation laboratory for evaluating dynamic traffic management systems. ASCE Journal of Transportation Engineering, 123(4), 283–289.
10. Myers, J. L., & Sadler, E. (1960). Effects of range of payoffs as a variable in risk taking. Journal of Experimental Psychology, 60(5), 306–309.
11. Busemeyer, J. R., & Townsend, J. T. (1993). Decision field theory: A dynamic cognitive approach to decision making. Psychological Review, 100(3), 432–459.
12. Erev, I., & Barron, G. (2005). On adaptation, maximization, and reinforcement learning among cognitive strategies. Psychological Review, 112(4), 912–931.
13. Avineri, E., & Prashker, J. N. (2003). Sensitivity to uncertainty: Need for a paradigm shift. Transportation Research Record: Journal of the Transportation Research Board, 1822, 90-98.
14. Denrell, J., & March, J. G. (2001). Adaptation as information restriction: The hot stove effect. Organizational Science, 12(5), 523–538.
15. Denrell, J. (2007). Adaptive learning and risk taking. Psychological Review, 114(1), 177–187.
16. Abdel-Aty, M. A. (1997). Using stated preferences data for studying the effect of advanced traffic information on drivers’ route choice. Transportation Research Part C, 5(1), 39–50.
17. Kaptein, N. A., Theeuwes, J., & van der Horst, R. (1996). Driving simulator validity: Some considerations. Transportation Research Record, 1550, 30-36.

Analogy for Understanding: Think of a driver choosing a route like a chef deciding whether to stick with a familiar recipe or try a new shortcut. If the chef is told the usual recipe will take 70 minutes longer than expected (numerical delay), they will almost certainly try the shortcut. However, an older, more experienced chef might switch to the shortcut even if the delay is small, because they remember too many times when a “small” delay turned into a kitchen disaster (the hot stove effect). Meanwhile, the 511 system is like a pre-shift briefing; although it provides the same information early, the chef often makes their final decision only when they actually see the “sold out” sign on their ingredients (the DMS) while they are already cooking.